Dr Hawkey points out that Somerville et al calculated that 200 deaths a vear are attributable to these drugs in Britain.' This is not so; the paper estimated that 200 deaths a vear from peptic ulcer haemorrhage occur due to non-steroidal antiinflammatory drugs. TFhis figure did not include deaths from perforation or sudden deaths occurring at home. Estimates as high as 3000-4000 deaths a year from peptic ulcer complications associated with use of non-steroidal anti-inflammatory drugs have been made." Clearly this would considerably alter the cost effectiveness of coprescribing an ulcer healing drug. Morbidity from peptic ulcers must also be taken into account. In a recent drug surveillance study based in Tayside 57 715 prescriptions for nonsteroidal anti-inflammatory drugs issued to 25 959 patients were analvsed by record linkage." Inpatient morbidity from upper gastrointestinal events was compared with that in community controls. Non-steroidal anti-inflammatory drug takers were shown to be more likely to develop gastrointestinal symptoms or disease. The relative risk for women was 1 67 and for men 1 28. Extrapolation from the data in this study indicates that non-steroidal anti-inflammatory drugs potentially cause 15 000 serious gastrointestinal events (perforation or haemorrhage) yearly in the United Kingdom. In addition many patients develop a simple gastric ulcer and considerably more have less severe gastroduodenal damage. There is now evidence that gastroduodenal damage can be prevented by coprescribing misoprostol," and it is therefore likely, but not proved, that the complications and mortality from peptic ulceration in users of non-steroidal antiinflammatory drugs can be reduced. Coprescribing an ulcer healing drug may be both beneficial to the patient and cost effective."

mortalitv are thus invalid. Because it has small numbers of younger patients the Tayside study leaves open the important question of whether elderly patients are particularly sensitive to non-steroidal anti-inflammatory drugs as opposed to particularly prone to ulceration. Although my article was not principally about individual drugs, Dr Shield points out that misoprostol can prevent gastric and duodenal erosions caused by ibuprofen. He points out that gastric erosions are more common than duodenal injury. However, as Graham and Smith remarked, "rhe clinician therefore should not be greatly concerned by the presence of gastric erosions associated with the acute administration of any drug given, because the extent and degree of such changes have no proven value for predicting the frequency and severity of these erosions, chronic ulcer or overt GI bleeding." On this basis most of the lesions portrayed in Dr Shield's dottogram are those with which the clinician "should not be greatly concerned." In the ibuprofen study there were 3/30 acute gastric ulcers and 1/30 acute duodenal ulcers on placebo with none on misoprostol. In the tolmetin study 13/30 patients on placebo, 4/30 patients on cimetidine, and 0/29 on misoprostol developed gastric ulcers while corresponding figures for duodenal ulceration were 3/30, 0/30, and 0/29. The reductions in tolmetin induced gastric ulceration by both cimetidine and by misoprostol were statistically significant, but any other comparisons of ulcer incidence in these studies were not. Moreover, it is not clear that these reductions are necessarily clinically important. The high incidence of lesions classified as ulcers supports my original argument that lesions classified as ulcers in these studies are unlikely to be as dangerous as ulcers not associated with non-steroidal anti-inflammatory drugs when A L BLOWAER viewed in the context of epidemiological studies Dcpartment ot' Surgery, such as the ones Dr Shield quotes. Hope Hospitail, Salford M6 SH ) I do not agree that protecting the stomach is more indeed less) important than protecting the (or 1 Hawkev CJ. Non-steroidal anti-inflammatory drugs and peptic duodenum. Likewise, in practical terms most tilcers. BrMAledJ7 1990;300:278-54. (3 Februarv.) clinicians will have to deal with non-steroidal anti2 Somerville K, Faulkener G, Langman MJS. Non-stcroidal antiinflammatorv drugs and blcedinig pcptic ulccr. Launci inflammatory drug takers who have presented with 1986;i:462-4. ulceration. No study has specifically investigated 3 Blower AL. Armstrong CIP. Sudden dcaths at horne from peptic this very important problem, but studies which ulceration-a hitherto unrecognised phenomenon. (Ga include patients with pre-existing ulcers probably 1986;27:A 1281. 4 Armstrong CP, Blower Al.. Nott-stcroidal anti-inflammatory have more relevance to the prescriber, rather than drtigs and tic lifc threateninig cotuplications of peptic uilceraless. tiott. Gut 1987;28:527-32. My article refrained from making recommenda5 Artnstrong CP, Whitclaw S. )eath trom undiagnosed pcptic uilcer complications: a cotttinuinig cltallttigc. Br.7 Surg 1988; tions about prophylaxis in view of the discrepan75:1112-4. cies of the data which it was written to highlight. 6 Blowcr AL, Armstrong CP. Ulcer perfOration in the elderly andl While there may be commercial pressures to wish noni-stcrotdal anti-inlamntiatory drugs. Lanicet 1986;i:97 1 7 (Cokal R. NSAIDs- should evcry prcscription carry a governto believe that there is a large and easily preventmettt hcalth warning. (Gu 1987;28:515-8. able problem in a conveniently identifiable group, 8 Beardon P'HD, Brown SX, McD)evitt )G. Gastrointestinal assessment of all available data still suggests that events itt patients prescribed ton-steroidal antiinflammatory these are genuinely open questions. It is important drtugs: a controlled study usiling rccord linkage in Tavside. QjMcd 1989;71:497-505. to identify questions which remain unanswered 9 (irahamI)Y, Agrawal NM9, Roth S. Prevention otf NSAII)because this highlights areas where continuing indtuced gastric ulcer with misoprostol: multicentrc double research is desirable. This should not be construed blind placcho-controlled triaill. Lantt 19XX;ii: 1277-81. as representing an implied commentary on the 10 Hililtani AL, Bloom BS. Econiomic effctts ot prophvlactic uise of tluisoprostol to prevsett gastric uilcer in patients taking noutmerits of competing commercial products. stertidal anti-inrflatutatory drugs. Arch Intern Med 1989;149: Mr Blower's points are in agreement with the 2116 1-5. essential arguments of my article-that estimates of risk are widely discrepant. His data were AUTHOR'S REPLY,-The Tayside study' was included in the meta-analvsis from which the published after my review was written but supports epidemiological estimates of risk quoted were the points I made. It generates an estimate of derived. relative risk of 2- 1(95 (ho confidence interval 1 * 77 to Gastroduodenal damage induced by non2 51), a figure rather lower than the average quoted steroidal anti-inflammatory drugs can undoubtin my article. Apparently higher rates of absolute edly be prevented by prostaglandins. Specific risk than those in case control studies of ulcer prophylaxis of lesions which are likely to bleed is bleeding arise because the International Classifica- less certain, while prevrention of complications has tion of Disease codings used would have included not been shown for any drug. Prophylaxis targeted bleeding which was trivial, incidental to admission, specifically at bleeding is as worth exploring as originating from parts of the gastrointestinal tract prophylaxis targeted at mucosal injury. In Mr Blower's study 26 patients had bleeding other than the stomach or duodenum or from lesions other than ulcers, as well as misdiagnoses of gastric ulcers and 69 had bleeding duodenal ulcersg gastrointestinal bleeding such as haemoptysis while in Graham et al's misoprostol trial of prophy(P H G Beardon, personal communication). laxis of acute lesions4 the vast majority had gastric lesions, two more sets of facts which do not add up. Calculations of absolute risk based on a 10WI

816

It would be a mistake to regard Graham et al's successful prophvlaxis of acute gastric lesions as a guide to prophylaxis of the chronic duodenal lesions which predominated in Mr Blower's studN. One way round this dilemma would be to postulate that non-steroidal anti-inflammatory drug induced haematemesis and melaena arise in whole or part because the drugs provoke bleeding in preexisting chronic ulcers. As prophylaxis of acute non-steroidal anti-inflammatory drug induced bleeding, manoeuvres which raise intragastric pH appear to be as successful as prostaglandins.' We have explored the possibility that normal clinical doses of the drug ethamsylate might reduce aspirin induced bleeding. Although the results failed to reach statistical significance, higher doses might yield better results and the general approach deserves further investigation. This is too important an area for premature conclusions to blight continuing research. C J IIAWKEY

l)Dpartrncnt of Therapetutics, Ulnicrsity Hospital, Nottingham NG7 2UH 1 Beardon 1'HG, Brown SV, McD)evitt 1)(i. (iastrointestinal events in patients prescrihed ttn-stnroidal anti-inflammatory drugs: A controlled study usittg rectrd lin kage in TaNside. Q 7 Mcd

1989;266:497-505.

2 (iraham DY, Smith JL. Aspirin and the stomach. Ann Intert Med

3 4

5 6

1986;104:390-8. Armstrong CP, Blower AL. Non-steroidal anti-inflammatory drugs and the life threatening complications of peptic ulceratiott. Gut 1987;28:527-32. (iraham DY, Agrawal NM, Roth SH. Presention of NSAII)induced gastric ulcer with misoprostol-a mtulticentre, double-blind, placebto-controlled trial. Lancet 1988;ii: 1277-80. Hawkey CJ, Somcr-ille KW, Marshall S. Prophylaxis of aspirin induLccd gastric mucosal bleeding with ranitidine. AlimentarIPhartatology and Therapeutics 1988;2:245-52. Stcin A, Daneshmend TK, Bhaskar NK, Hawkey CJ. Failure of cthatosvlate to reduce aspiriil induced gastric mucosal bleeding in

hutnatts. Brj Clin Pharmnacol 1989;28:109-12.

Electrical arcing and contact lenses SIR,-Welding arcs and high voltage electrical flashovers are powerful sources of light of many wavelengths. They may damage the eyes of anyone, especially if suitable screens, goggles, or glasses are not used. Arc eye or welder's flash (keratitis photoelectrica) is caused by damage to the corneal epithelium, which absorbs light in the ultraviolet wavelengths. After a latent period severe pain ensues but the epithelium soon regenerates and there is usually no residual damage. Most contact lenses do not protect against arc eye. Furthermore, they absorb energy from parts of the ultraviolet and infrared spectrums, and the resultant rise in temperature may cause them to adhere temporarily to the eye. Lovsund et al, using thick hydrogel lenses, showed that a rise of 81 OC above the temperature of the surrounding air was produced in five minutes. The effect could be minimised by the use of an industrial screen. In actual lens wear tear flow and blinking could be expected to moderate the rises in temperature. There have been reports of severe electric arc damage to workers' eyes in many countries since the 1960s. One much repeated story arose after a shipyard incident in the United States in 1977, though the official inquiry established contact lens overwear as the cause. A common variant of this tale appeared in a national newspaper in May 1989.2 It was comprehensively denied bv the Health and Safety Executive two days later. T he story will undoubtedly continue to surface from time to time-as Minerva has discovered'-sometimes innocently propagated by overcredulous industrial health advisers. Contact lens wearers who are exposed to electric arcs should use the same recommended eve protection as those who do not use lenses and should blink frequently. If aphakic they should wear

BMJ VOLUME 300

24 MARCH 1990

ultraviolet contact lenses that absorb ultraviolet

light so as to protect their retinas from damage. R J BUCKLEY

Moorfields Eye Hospital, London ECIV 2PD 1 Lovsund P, Nilsson SEG, Lindh H, Oberg PA. Temperature changes in contact lenses in connection with radiation from welding arcs. Scandj Work Envtron Health 1979;5:271-9. 2 Leadbetter C. Welders are told to take off contact lenses. Financial Times 1989 May 22. 3 Health and Safety Executive. Time to focus on some hocus pocus. News release E62:89. London: Health and Safety Executive, 1989. 4 Minerva. Views. BrMedJ 1990;300:620. (3 March.)

Fibromuscular dysplasia of renal arteries and acute loin pain SIR,-In response to the comments of Dr M J Duddy and colleagues on our paper2 we wish to re-emphasise that delay in diagnosis of an acute -tenovascular emergency may be disastrous. Prolonged discussion of the relative incidence of the causes of loin pain and of a non-functioning kidney is futile, but in any case we would strongly dispute that ureteric obstruction is the commonest cause. We do not know how often renovascular emergencies occur, but non-functioning or contracted kidneys are seen often enough in intravenous urograms of patients without symptoms to suggest a fair measure of unrecognised acute disease. We regret the implication that we do not take a history or examine our patients before imposing tests on them. We strongly disagree that intravenous urography has been displaced by ultrasonography and plain abdominal x ray films in patients with loin pain. There is no certainty that an -opacity is a calculus, and pyuria and bacteriuria can coexist with stone disease. Furthermore, as we emphasised, obstruction is not excluded by an undilated pelvicaliceal system on ultrasonography. Intravenous urography is also necessary to confirm normal contralateral function. We still maintain that very poor or absent function shown on urography with an undilated pelvicaliceal system on ultrasonography should lead to angiography. Case 1 shows the catastrophic effects of delay while other tests were being done. We did not discuss further investigation of the non-functioning hydronephrotic kidney because it was not relevant. In any case we would disagree that antegrade pyelography should necessarily be done. The level of obstruction may be clinically obvious, and, particularly in the presence of overwhelming malignancy, it may 'be entirely inappropriate to relieve obstruction by percutaneous nephrostomy. Finally, we agree that where there is a small, poorly functioning kidney another cause for the patient's symptoms should be sought; no further investigations of the kidney are indicated. M C BISHOP

A R MANHIRE

S J STINCHCOMBE Departments of Radiology and Urology, City and University Hospitals, Nottingham NG5 IPB

R H J GREGSON

I Duddy MJ, Bradley SA, Chapman S. Fibromuscular dysplasia of renal arteries and acute loin pain. Br Med J 1990;300:469.

V 17 February.) 2 Stinchcombe SJ, Manhire AR, Bishop MC, Gregson RHJ. Fibromuscular dysplasia of renal arteries: a neglected cause of acutc loin pain. Br MedJ7 1990;300:183-5. (20 January.)

Bovine spongiform encephalopathy SIR,-Professor W B Matthews gave an interesting account of the controversy that exists over the relation between bovine spongiform encephalopathy and scrapie in cattle and sheep, and kuru and

I

(l.MI 300

24 MARCII 1990

Creutzfeldt-Jakob disease in humans.' We were, however, concerned at his point that all suspected cases of Creutzfeldt-Jakob disease "must be seen by a neurologist who is familiar with the disease," and we wonder how often this is practical. In a recent review of the disease Knight listed various symptoms and their frequency of occurrence, with dementia as an invariable feature of all forms, often as an initial symptom.2 The variable presentations include sleep disturbance, weight loss, and non-specific headaches, possibly suggesting a depressive picture; indeed these features may exist in isolation for a fairly prolonged period. In addition, a floridly psychotic picture may appear, with visual and auditory hallucinations accompanied by a rapidly developing dementia.3 The variety of these symptoms in presentation, accompanied by the commonly held view of dementias, often results in patients being referred to a psychiatrist rather than to a neurologist before neurological symptoms develop. It is well recognised that electroencephalography in Creutzfeldt-Jakob disease is a helpful investigation whereas few others are: there is a characteristic pattern of generalised biphasic or triphasic complexes.2 For this reason, we have found it useful to obtain a baseline electroencephalogram for all new patients presenting with a dementing illness, repeating the investigation if there is any substantial change in their mental or neurological state. Should there be reason to suggest Creutzfeldt-Jakob disease, the opinion of a neurologist can then be sought, although, with an annual incidence estimated at 0 49/million in England and Wales (population about 50 million) -about 25 expected new cases annually-few practitioners can have access to a neurologist who is "familiar with the disease." Although Creutzfeldt-Jakob disease is a rare disease of uncertain aetiology, an index of suspicion should exist amongst all practitioners who deal with cases of dementia. We would recommend electroencephalography as a baseline investigation and would question whether the estimates of incidence can be considered valid for a disease that can present in so many different specialties. Given that this figure may not be accurate, what price epidemiological studies of the relation between bovine spongiform encephalopathy and Creutzfeldt-Jakob disease? CHRISTOPHER FEAR MANIKKARASA DEVAKUMAR Ysbyty Gwynedd, Bangor, Gwyniedd LL5 1 2PW I Matthews WB. Bovine spongiform encephalopathy. Br Med J

1990;300:912-3. (17 February.) 2 Knight R. Creutzfeldt-Jakob disease. Br7 Hosp Med 1989;41: 165-7 1. 3 Bethan P. Creutzfeldt-Jakob disease. BrMedJ 1982;284: 1658-9.

Judicial review of the NHS SIR,-After much discussion among ourselves and consultation with our legal advisers we have decided not to appeal against the High Court judgment of Lord Justice Woolf and Mr Justice Pill. In their judgment the 1977 NHS Act offers the secretary of state virtually unrestricted spending powers, extending even to structures and measures as yet without a basis in law. Their conclusions rested heavily on the interpretation of a sentence and the significance of a comma. Our initial inclination was strongly to appeal against this judgment. However, even had the appeal succeeded it would almost certainly have been further challenged in the House of Lords. Much money would have been spent, much time passed before a final verdict late this year or early next-an increasingly academic exercise, with the real issues of authoritarian manipulation not coming through in court. It can give the minister little satisfaction to know

that he has scored a technical victory over the great body of consultants who offer us moral and financial support. Let him make no mistake; this was not the action of some group offended by an attack on professional privilege. It was the voice of the great majority of doctors, nurses, those in the professions ancillary to medicine, and the public itself, protesting in the only potentially effective way we could see against a bill that we feel threatens and demeans the true purposes of medicine. That has not changed. It was a protest too against the methods and morality of the boardroom that are replacing the processes of democratic consent. Ministers should be servants of the public, not its masters. They should be the custodians of the NHS, not its owners. It is not too late for the secretary of state and those to whom he answers to redress the wrongs they seem so firmly determined to impose on profession and public, to accept reasoned amendments, and to respond to sincere concerns by modifying the terms of the bill. The conviction that the government is right and that we must be brought to see how wrong we are has more the flavour of the eastern European "democracies" now passing into history than of Britain with its record of hard fought freedoms. We have decided not to appeal, but let that not be thought to mean that we accept what is being done or the manner of its doing. The profession and the public resent their exclusion from contributing their ideas to the shape of our future medical services. This unity of purpose and concern will be carried forward in the NHS Support Federation, a new alliance of the public and the professions, which is determined to preserve and promote the great purposes of the NHS, whatever the government of the day. H KEEN

NHS Support Group for Judicial Review, Guy's Hospital, London SE l 9RT

How to pay your way through medical school SIR,-In response to Mr A P Armstrong's article I would like to say that, in London at least, drug trials contribute vastly more to student incomes than' charity. By qualification, it is possible for a student to have taken fi blockers and H2 antagonists, antipsychotics, and antidepressants; to have undergone nasogastric intubation, intravenous cannulation, and placement of an indwelling catheter; and to have had a liver biopsy and repeated bone marrow aspirations -and all against a background of weekly semen donation. Female students are barred from many of these trials, but if they have premenstrual tension or menorrhagia they can take vitamin B-6, evening primrose oil, mefenamic acid, or progesterone. In the event of student loans many students will turn to drug trials as a source of tax free income. No trial, no matter how suspect, will be refusable. Students with moderate asthma will not seek treatment from their general practitioners as such treatment would bar them from (i blocker or atopy trials. Female students will all have to claim premenstrual tension of psychotic proportions and gross menorrhagia. A situation can be envisaged in which dangerous trials can recruit subjects, student illnesses go untreated, and trial subjects do not have the conditions relevant to the trial. Economic coercion of this sort would serve not the student, not the drug company, and not the public but would benefit only the Treasury. JAMES BARRETT London N19 5SE 1 Armstrong AP. How to do it: pay your way through medical school. BrM11edJ 1990;300:453-4. (17 February.)

817

Electric arcing and contact lenses.

Dr Hawkey points out that Somerville et al calculated that 200 deaths a vear are attributable to these drugs in Britain.' This is not so; the paper es...
642KB Sizes 0 Downloads 0 Views